Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A decoding device comprising: circuitry configured to receive coded data and conversion information, the coded data pertaining to an image having luminance in a first dynamic range and the conversion information pertaining to a conversion of dynamic range of the luminance of the image from the first dynamic range into a second dynamic range; and decode the received coded data so as to generate the image, wherein the conversion uses a knee function that is defined by at least one knee point, and wherein the conversion information includes unconverted luminance range information indicating a permillage of peak luminance level for the image prior to the conversion, relative to a normalized luminance level, unconverted display luminance information indicating an expected display brightness of peak luminance level for the image prior to the conversion, converted luminance range information indicating a permillage of peak luminance level for the image after the conversion relative to the normalized luminance level, and converted display luminance information indicating an expected display brightness of peak luminance level for the image after the conversion.
The invention relates to image decoding systems that handle dynamic range conversion. The problem addressed is efficiently decoding images with luminance values in a high dynamic range (HDR) while ensuring accurate display output in a lower dynamic range (LDR) or another target range. The solution involves a decoding device with circuitry that processes coded image data and conversion metadata. The coded data represents an image with luminance in an original dynamic range, while the conversion information specifies how to map this luminance to a target dynamic range using a knee function—a nonlinear transformation defined by at least one knee point. The conversion metadata includes four key parameters: (1) unconverted luminance range information, indicating the original peak luminance as a fraction of a normalized level (e.g., 1000 parts per million), (2) unconverted display luminance information, specifying the expected display brightness for the original peak luminance, (3) converted luminance range information, indicating the target peak luminance as a fraction of the normalized level after conversion, and (4) converted display luminance information, specifying the expected display brightness for the converted peak luminance. This approach ensures precise dynamic range adaptation while maintaining visual fidelity. The circuitry decodes the coded data and applies the knee function using the provided metadata to generate the final image.
2. The decoding device according to claim 1 , wherein the conversion uses the knee function to map the dynamic range of the luminance of the image from the first dynamic range into the second dynamic range.
This invention relates to image processing, specifically to a decoding device that converts the dynamic range of an image from a first dynamic range to a second dynamic range. The problem addressed is the need to accurately map luminance values between different dynamic ranges, particularly in high dynamic range (HDR) imaging, where preserving perceptual quality is critical. The decoding device includes a conversion module that applies a knee function to adjust the luminance values of the image. The knee function is a nonlinear transformation that compresses or expands the dynamic range while maintaining visual fidelity. It ensures that highlights and shadows are preserved without excessive clipping or loss of detail. The function can be parameterized to adapt to different input and output dynamic ranges, allowing flexible conversion between various HDR and standard dynamic range (SDR) formats. The knee function typically involves a piecewise linear or sigmoidal curve that smoothly transitions between linear and logarithmic regions. This approach prevents abrupt changes in brightness and ensures a natural appearance. The device may also include additional processing steps, such as tone mapping or color correction, to further enhance the image quality after dynamic range conversion. This technology is useful in applications like HDR video playback, professional video editing, and display systems where accurate luminance mapping is essential for high-quality visual output.
3. The decoding device according to claim 2 , wherein the conversion information includes pre-conversion information and post-conversion information, wherein the pre-conversion information indicates a range of luminance which is a knee function target in the first dynamic range, and wherein the post-conversion information indicates a range of luminance in the second dynamic range that corresponds to the range of luminance indicated, by the pre-conversion information.
A decoding device processes video data with different dynamic ranges, particularly for high dynamic range (HDR) to standard dynamic range (SDR) conversion. The device converts luminance values between a first dynamic range (e.g., HDR) and a second dynamic range (e.g., SDR) using conversion information. This conversion information includes pre-conversion and post-conversion data. The pre-conversion information defines a luminance range in the first dynamic range that is targeted by a knee function, which adjusts luminance values to avoid clipping or excessive brightness. The post-conversion information specifies the corresponding luminance range in the second dynamic range, ensuring accurate mapping between the two ranges. This allows the device to preserve visual quality during dynamic range conversion, addressing challenges in maintaining perceptual fidelity when converting between HDR and SDR content. The conversion process ensures that specific luminance ranges are accurately transformed, preventing loss of detail in bright or dark regions. The device may also include a memory to store the conversion information and a processor to apply the conversion during decoding. This technology is useful in video playback systems where dynamic range conversion is required for compatibility with different display devices.
4. The decoding device according to claim 3 , wherein the pre-conversion information indicates the range of luminance which is converted by knee function at a same conversion ratio as a conversion range of the first dynamic range.
This invention relates to a decoding device for processing video signals, specifically addressing the challenge of accurately converting luminance values between different dynamic ranges while preserving visual quality. The device includes a pre-conversion information acquisition unit that retrieves pre-conversion information, which specifies the luminance range that undergoes conversion using a knee function. The knee function is a nonlinear transformation applied to adjust luminance values, particularly in high dynamic range (HDR) to standard dynamic range (SDR) conversion. The pre-conversion information indicates the specific luminance range where the knee function applies the same conversion ratio as the original dynamic range, ensuring smooth transitions and avoiding artifacts. The device also includes a conversion unit that processes the luminance values based on this information, applying the knee function only to the designated range while maintaining the original conversion ratio for that segment. This approach prevents abrupt changes in brightness and improves visual consistency. The invention is particularly useful in video decoding systems where dynamic range conversion is required, such as in HDR to SDR conversion for display compatibility. By dynamically adjusting the conversion process based on pre-conversion information, the device ensures accurate and visually pleasing results.
5. The decoding device according to claim 1 , wherein the conversion uses the knee function which is defined by a plurality of knee points.
A decoding device processes encoded data by converting it into a target format. The conversion employs a knee function, which is defined by multiple knee points. These knee points are specific points on a curve that adjust the conversion process to achieve desired output characteristics. The knee function allows for nonlinear adjustments, enabling precise control over the conversion process. The device may include an input interface to receive the encoded data and an output interface to provide the converted data. The knee function can be dynamically adjusted based on input parameters or predefined settings to optimize the conversion for different types of encoded data. This approach ensures accurate and efficient decoding while maintaining flexibility in handling various input formats. The use of multiple knee points allows for fine-tuning the conversion process, improving the quality and reliability of the decoded output. The device may also include processing circuitry to execute the conversion and memory to store the knee function parameters. This method enhances the performance of decoding systems by providing a flexible and adaptable conversion mechanism.
6. The decoding device according to claim 5 , wherein the conversion information includes a plurality of pairs of the pre-conversion information and the post-conversion information.
A decoding device is designed to process encoded data by converting pre-conversion information into post-conversion information using conversion information. The conversion information consists of multiple pairs, where each pair links specific pre-conversion information to its corresponding post-conversion information. This allows the device to accurately reconstruct or decode the original data by applying the appropriate conversion rules stored in these pairs. The device may be part of a larger system that handles data encoding and decoding, ensuring that the original information is accurately retrieved from the encoded format. The use of multiple pairs in the conversion information enables flexibility in handling different types of encoded data, improving the device's ability to decode various formats efficiently. This approach is particularly useful in applications where data must be decoded from a compressed or encrypted form, ensuring reliable and accurate reconstruction of the original information.
7. The decoding device according to claim 1 , wherein the conversion uses the knee function by mapping the dynamic range of the luminance of the image from the first dynamic range into the second dynamic range, and a plurality of adjacent segments of the first dynamic range of the luminance are mapped to a corresponding plurality of adjacent segments of the second dynamic range of the luminance based on boundaries between adjacent segments defined by a plurality of knee points.
This invention relates to image processing, specifically to a decoding device that converts the dynamic range of an image from a first dynamic range to a second dynamic range using a knee function. The problem addressed is the need for efficient and accurate dynamic range conversion, particularly in scenarios where different segments of the luminance range require distinct mapping to preserve image quality. The decoding device includes a conversion unit that applies a knee function to map the luminance values of an input image from the first dynamic range to the second dynamic range. The knee function divides the first dynamic range into multiple adjacent segments, each defined by knee points that serve as boundaries. These segments are then mapped to corresponding adjacent segments in the second dynamic range. This approach ensures that different portions of the luminance range are adjusted appropriately, maintaining visual fidelity. The knee function allows for non-linear adjustments, which are particularly useful in high dynamic range (HDR) to standard dynamic range (SDR) conversions or vice versa. The method ensures smooth transitions between segments, preventing abrupt changes in brightness that could degrade image quality. The invention is applicable in various display technologies, including HDR displays and tone-mapping applications.
8. The decoding device according to claim 2 , wherein the conversion uses the knee function by mapping the dynamic range of the luminance of the image from the first dynamic range into the second dynamic range at a first conversion ratio to a point defined by a first knee point of the at least one knee point and at a second, conversion ratio from the point defined by the first knee point.
This invention relates to image processing, specifically to a decoding device that converts the dynamic range of an image from a first dynamic range to a second dynamic range using a knee function. The knee function maps the luminance values of the image, adjusting the conversion ratio at specific points to optimize the dynamic range transformation. The conversion process involves applying a first conversion ratio up to a first knee point and then switching to a second conversion ratio beyond that point. This approach allows for precise control over how luminance values are adjusted, ensuring that details in both bright and dark regions of the image are preserved. The knee function is particularly useful in high dynamic range (HDR) imaging, where maintaining visual fidelity across a wide range of luminance levels is critical. By dynamically adjusting the conversion ratio at predefined knee points, the device avoids excessive compression or expansion of luminance values, resulting in a more natural and accurate representation of the original image. This technique is applicable in various imaging systems, including displays, cameras, and video processing pipelines, where dynamic range adaptation is necessary.
9. The decoding device according to claim 1 , wherein the knee function is specified by a Supplemental Enhancement Information (SEI) message.
A decoding device processes video data encoded with a knee function, which adjusts the dynamic range of pixel values to improve visual quality. The knee function is defined by a Supplemental Enhancement Information (SEI) message, a metadata structure in video coding standards like H.264/AVC or HEVC. This SEI message provides parameters that control the knee function's shape, such as input and output ranges, slope, and breakpoints, allowing flexible adaptation to different display environments. The device applies the knee function during decoding to map input pixel values to output values, ensuring optimal brightness and contrast for the target display. The SEI message may also include flags or indicators to enable or disable the knee function dynamically. This approach enhances compatibility with various displays and improves the viewing experience by preserving details in bright and dark regions of the image. The knee function can be applied to the entire frame or selectively to specific regions, as indicated by additional metadata in the SEI message. This method ensures that the decoded video maintains high quality across different display devices and lighting conditions.
10. The decoding device according to claim 9 , wherein the SEI message includes a setting of a knee_function_id.
A decoding device processes video data encoded with a knee function, which adjusts pixel values to improve visual quality. The device extracts a Supplemental Enhancement Information (SEI) message from the encoded video data, where the SEI message contains a knee_function_id setting. This identifier specifies the type of knee function applied during encoding, allowing the decoder to correctly interpret and apply the inverse knee function during decoding. The knee function typically modifies pixel values in a nonlinear manner to enhance dynamic range or contrast, and the knee_function_id ensures compatibility between encoder and decoder settings. The device uses this identifier to select the appropriate inverse knee function parameters, restoring the original pixel values accurately. This approach prevents mismatches between encoding and decoding processes, ensuring consistent visual quality. The SEI message may also include additional parameters defining the knee function's characteristics, such as control points or curve shapes, which the decoder uses to reconstruct the inverse transformation. This method is particularly useful in high dynamic range (HDR) video processing, where precise tone mapping is critical. The decoding device applies the inverse knee function to decoded pixel values, correcting any distortions introduced during encoding. This ensures that the final output video maintains the intended visual quality and dynamic range.
11. A decoding method of causing a decoding device to perform: receiving coded data and conversion information, the coded data pertaining to an image having luminance in a first dynamic range and the conversion information pertaining to a conversion of dynamic range of the luminance of the image from the first dynamic range into a second dynamic range; and decoding the received coded data so as to generate the image, wherein the conversion uses a knee function that is defined by at least one knee point, and wherein the conversion information includes unconverted luminance range information indicating a permillage of peak luminance level for the image prior to the conversion relative to a normalized luminance level, unconverted display luminance information indicating an expected display brightness of peak luminance level for the image prior to the conversion, converted luminance range information indicating a permillage of peak luminance level for the image after the conversion relative to the normalized luminance level, and converted display luminance information indicating an expected display brightness of peak luminance level for the image after the conversion.
This invention relates to image decoding techniques for handling dynamic range conversion in video processing. The problem addressed is the need to accurately decode and display images that have undergone dynamic range adjustments, ensuring proper brightness representation across different display environments. The method involves receiving coded image data and associated conversion information, where the image data represents an image with luminance in a first dynamic range, and the conversion information describes a transformation of that luminance into a second dynamic range. The decoding process generates the image while applying a knee function defined by at least one knee point to perform the dynamic range conversion. The conversion information includes multiple parameters: unconverted luminance range information specifying the permillage of peak luminance before conversion relative to a normalized level, unconverted display luminance information indicating the expected display brightness of peak luminance before conversion, converted luminance range information specifying the permillage of peak luminance after conversion relative to the normalized level, and converted display luminance information indicating the expected display brightness of peak luminance after conversion. These parameters enable precise reconstruction of the original and converted luminance values, ensuring accurate brightness representation during decoding and display. The method ensures compatibility between different dynamic ranges while maintaining visual fidelity.
12. A non-transitory computer-readable medium having stored thereon coded data and conversion information, the coded data pertaining to an image having luminance in a first dynamic range and the conversion information pertaining to a conversion of dynamic range of the luminance of the image from the first dynamic range into a second dynamic range, wherein when a decoding device decodes the coded data, the decoding device is caused to generate the image based on the decoded data, and also caused to convert, using a knee function that is defined by at least one knee point and based on the conversion information, the dynamic range of the luminance of the image from the first dynamic range into a second dynamic range, and wherein the conversion information includes unconverted luminance range information indicating a permillage of peak luminance level for the image prior to the conversion relative to a normalized luminance level, unconverted display luminance information indicating an expected display brightness of peak luminance level for the image prior to the conversion, converted luminance range information indicating a permillage of peak luminance level for the image after the conversion relative to the normalized luminance level, and converted display luminance information indicating an expected display brightness of peak luminance level for the image after the conversion.
This invention relates to dynamic range conversion in image processing, specifically for encoding and decoding images with different luminance dynamic ranges. The problem addressed is efficiently representing high dynamic range (HDR) images in a way that allows accurate conversion to lower dynamic ranges (e.g., standard dynamic range, SDR) while preserving visual quality. The solution involves storing coded image data alongside conversion information that enables precise dynamic range adjustment during decoding. The system encodes an image with luminance in a first dynamic range (e.g., HDR) and includes conversion information to transform it into a second dynamic range (e.g., SDR). The conversion uses a knee function defined by at least one knee point, which non-linearly adjusts luminance values. The conversion information includes: 1. Unconverted luminance range information: a permillage (parts per thousand) of peak luminance before conversion relative to a normalized level. 2. Unconverted display luminance information: expected display brightness of peak luminance before conversion. 3. Converted luminance range information: permillage of peak luminance after conversion relative to the normalized level. 4. Converted display luminance information: expected display brightness of peak luminance after conversion. When decoded, the image is reconstructed and its dynamic range is converted using the knee function and the provided conversion parameters. This ensures accurate and visually consistent transformation between dynamic ranges while maintaining the original image's perceptual quality. The approach is particularly useful for HDR content distribution to SDR displays.
Unknown
March 17, 2020
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.